fix: do not delete cache on destroy

[gastonfournier]

Description

Fix UnleashClient.destroy() behavior so it no longer deletes the default SDK FileCache on shutdown.

Previously, shutdown always called cache.destroy(). For the default FileCache, this could remove shared on-disk cache data and break other live processes using the same app_name cache path (reported in #390). The fix preserves shutdown behavior for connector/scheduler/metrics while skipping cache deletion for default FileCache. Custom cache implementations still receive destroy() as before.

This is a targeted mitigation for the teardown race in #390. It does not fully address broader multiprocessing limitations of fcache discussed in #303.

Fixes #390

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Added/updated unit tests:

• test_destroy_skips_default_file_cache_destroy
• test_destroy_calls_custom_cache_destroy

Also updated cache fixture isolation in unit tests (tmpdir.strpath) to avoid cross-test cache leakage now that default FileCache is not deleted during client shutdown.

• Unit tests
• Spec Tests
• Integration tests / Manual Tests

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules

[sighphyre]

I think we can just remove the cache destruction code here but this is probably safe

[Copilot]

Pull request overview

Adjusts UnleashClient.destroy() behavior so destroying a client no longer wipes the on-disk cache, and updates unit tests to validate the new cache lifecycle expectations.

Changes:

• Removed cache teardown from UnleashClient.destroy().
• Updated the unit-test FileCache fixture to use a per-test temp directory since caches are no longer cleared on destroy.
• Added tests asserting destroy() skips FileCache.destroy() but still calls destroy() on custom cache implementations.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
UnleashClient/__init__.py	Changes client teardown logic by removing cache destruction during destroy().
tests/unit_tests/test_client.py	Updates cache fixture path handling and adds tests covering the new destroy() cache behavior.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

UnleashClient.destroy() no longer calls self.cache.destroy(), which breaks the expected lifecycle contract for custom caches (and contradicts test_destroy_calls_custom_cache_destroy). Consider restoring cache teardown for non-FileCache caches (e.g., if self.cache and not isinstance(self.cache, FileCache): ...) so custom implementations can release resources while keeping FileCache on disk.

Suggested change

	# Tear down custom caches while keeping FileCache contents on disk.
	cache = getattr(self, "cache", None)
	if cache and not isinstance(cache, FileCache):
	try:
	cache.destroy()
	except Exception as exc:
	LOGGER.warning("Exception during cache teardown: %s", exc)

[Copilot]

destroy()'s docstring mentions deleting the cache, but cache teardown is no longer performed here. Please update the docstring to match the new behavior (e.g., preserve FileCache on disk while optionally destroying custom caches).